Tracking your transit damage rate: what to measure and what it tells you

Most online sellers know last week's sales to the rupee. Ask them how many orders arrived broken, and the answer is usually "a few" or "not many".

"A few" cannot be fixed, because nobody knows which few.

Transit damage is the reason people buy our cushioning. Even so, I would put a simple damage log ahead of any cushioning we could sell you, because the log tells you where cushioning is actually needed, and where it is not.

This guide shows how to set up that log in a spreadsheet, what to record, and how to read it so that it points to an action. It does not put a rupee value on each broken order. That is a separate exercise. Here we count first, and then we look for the pattern.

What counts as a damaged order?

Write your definition down before you count anything. If two people in the team count differently, the number moves for reasons that have nothing to do with packing.

The definition I suggest is this. A damaged order is a delivered order where the product, or its own inner pack, reached the customer broken, cracked, dented, leaking, scratched or not working, because of something that happened after it left your dock.

Then write down what it is NOT, because these get mixed in all the time:

1. Lost parcels. They belong to the courier, so count them in a separate column.
2. Wrong or missing items. That is a picking error.
3. Defects that were there before packing. A product that never worked was not damaged on the road.
4. A dented outer box with the product fine. Keep this as a separate "box only" count.

I would watch the fourth one closely.

A box-only complaint costs you nothing today. But when that count rises, the outer pack is taking more punishment than before, and broken products can follow if nothing changes.

The formula, and why the dispatch week matters

Damage rate = damaged orders ÷ delivered orders.

I like to write it per 1,000 orders rather than as a percentage. A packing team remembers "6 per 1,000" more easily than "0.6%". (Those figures only show the format. They are not a target.)

Count both numbers by the week the orders were dispatched, not the week the complaint arrived. A parcel packed on Monday may reach the customer on Thursday, and the complaint may come a week after that. If you count complaints by the week they come in, one bad week of packing gets spread over the next three weeks, and you can never match a spike to its cause.

This means the latest week is never final.

Mark each week "open" until the return window for its orders has closed, and only then lock it.

Where does the damage data come from?

No single source catches everything. Use all of these and join them by order ID:

1. Customer complaints on phone, email and WhatsApp.
2. Marketplace returns where the reason given is "damaged" or "defective". Read the customer's comment too, because people do not always pick the right reason from the list.
3. Claims you raised with couriers.
4. Free replacements your team sent out.
5. RTO parcels. Open every parcel that comes back and check it. An RTO parcel has made the trip twice, so give RTO damage its own column.
6. Ratings and reviews that mention broken, leaking or crushed products.

Then remove the duplicates. One broken order can arrive as a complaint, then a return, then a one-star review. It is still one damaged order.

Expect the number to be on the low side. Many customers do not complain about a small chip or a dented tin, especially on a low-priced item. They simply do not order again.

So treat your logged rate as the floor, not the full picture.

What should you record for each damaged order?

Keep the columns few, so the person who fills them actually fills them. These are the ones I would not drop:

The last three columns are where the learning sits. A broken product in a perfect box tells a very different story from a broken product in a box with a hole in its side. Our guide to why products break in transit explains how to read those photos and match them to a cause.

Ask every complaining customer for two photos before they throw anything away: the outer box as it arrived, and the product inside it. Put that request into your standard reply.

Most of the evidence goes into the dustbin within an hour of opening.

Split the number before you act on it

One overall rate tells you whether things are getting better or worse. It does not tell you what to do. For that, split it:

1. By SKU, to find the products that break most.
2. By pack spec, when the same product goes out in more than one pack.
3. By courier, and by surface or air service.
4. By destination, at least by state, and by metro versus smaller towns.
5. By packer, station or shift.
6. By month, so the monsoon and the festival rush show up.

If you only have time for one split, I would start with SKU. It is the split most likely to point at a pack you can change this week.

Every split needs its own denominator. Let us say a glass jar had 3 damaged orders out of 40 dispatched, and a steel bottle had 10 out of 5,000. The bottle has more complaints, but the jar is the problem. So export your dispatch counts by the same splits: SKU, courier and destination.

A simple weekly sheet

You do not need software for this. A shared spreadsheet with three tabs is enough to start.

1. Dispatch tab. Paste the weekly export from your order system: order ID, dispatch date, SKUs, courier and pin code.
2. Damage log tab. One row per damaged order, with the columns above.
3. Summary tab. One row per dispatch week, like this:

I would give the sheet one owner. It works best with someone from the packing side, not accounts, because they can recognise a pack spec from a customer's photo.

Then spend fifteen minutes on it every week with the packing lead (and whoever answers complaints). Look at three things only: the trend, the biggest split, and whether last week's action changed anything.

What the pattern tells you

This is where the log starts paying for the effort. The pattern usually points to one of a few places:

1. One SKU is high with every courier and in every region. The pack for that SKU is the problem. Test a better pack before you change anything else. A simple drop test in your own warehouse is enough to compare two options.
2. One courier is high across many SKUs. Look at the outer-box photos. If boxes arrive crushed or punctured with that courier and not with others, take the photos to them, and think about moving your fragile SKUs.
3. One destination is high. Long routes pass through more hubs, which means more handling. You may need a stronger pack for that lane only, not for every order.
4. One packer or shift is high. That is a training matter, not a material matter. Show the team the photos, not just the number.
5. Crushed boxes rise in July and August. Damp weather weakens corrugated boxes, so check the box grade and how your empty cartons are stored.
6. Leaks rise on one product. Look at the cap, the seal and the way the bottle stands in the box before you add cushioning.

I will be honest about one bias here. We make cushioning, so it would suit us if the answer were always "use more of it". A good log often says something else. Many times the fix is a better-fitting box, a different courier for one lane, or ten minutes of training.

Follow what the log says, even when the answer is not a new material.

After you change something

Change one thing at a time, and write the date of the change in the summary tab. Then compare the same SKUs over the same number of weeks, before and after.

Be patient with small numbers. If an SKU ships 30 orders a week, a single breakage moves its rate a lot. It is like a blood sugar test. One reading taken after a wedding dinner tells you very little. The readings across a month tell you a lot.

For low-volume SKUs, look at a rolling four weeks and read the photos, rather than reacting to one bad week.

What is a good damage rate?

Everyone wants a benchmark. I will not give you one.

A rate that is fine for steel kitchen tools would be a disaster for glass perfume bottles. The route and the price of the product change it again. A number borrowed from another business can make you relax when you should not, or spend on packing you do not need.

Your real benchmark is your own last twelve weeks. If the line goes down and stays down, you are doing the right things.

If your log keeps pointing at one product and you are not sure what to change, send us the SKU, its weight and two damage photos. We will look at them with you and suggest a pack you can test against your current one.